Method and apparatus for assessing the quality of spectral images
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
G06K-009/46
출원번호
US-0695797
(2000-10-23)
발명자
/ 주소
Sweet, James N.
출원인 / 주소
BAE Systems Mission Solutions Inc.
대리인 / 주소
Cooley Godward LLP
인용정보
피인용 횟수 :
11인용 특허 :
8
초록▼
A method is disclosed herein for evaluating quality of an image. The disclosed method contemplates receiving a spectral image and extracting a plurality of pixels therefrom. The plurality of pixels are converted into a plurality of spectral vectors, wherein each element in each spectral vector repre
A method is disclosed herein for evaluating quality of an image. The disclosed method contemplates receiving a spectral image and extracting a plurality of pixels therefrom. The plurality of pixels are converted into a plurality of spectral vectors, wherein each element in each spectral vector represents a property of a respective one of N spectral bands. The plurality of spectral vectors are then categorized into a set of M classes. The method further includes the step of computing a mean vector for each of the M classes based upon the spectral vectors associated therewith. Next, spectral similarity values between pairs of the mean vectors are computed. The distribution of these spectral similarity values may then be analyzed in order to obtain information relevant to image quality.
대표청구항▼
1. A method for evaluating quality of an image, comprising:receiving a spectral image;extracting a plurality of pixels from the spectral image;converting the plurality of pixels into a plurality of spectral vectors, each element in each spectral vector representing a property of a respective one of
1. A method for evaluating quality of an image, comprising:receiving a spectral image;extracting a plurality of pixels from the spectral image;converting the plurality of pixels into a plurality of spectral vectors, each element in each spectral vector representing a property of a respective one of a number (N) of spectral bands;categorizing said plurality of spectral vectors into a set of M classes;computing a mean vector for each of said M classes based upon the spectral vectors associated therewith;computing spectral similarity values between pairs of said mean vectors; andanalyzing said spectral similarity values in order to obtain information relevant to said quality of said image. 2. The method of claim 1 wherein said categorizing includes computing spectral similarity values between a seed one of said plurality of spectral vectors and each of a set of unprocessed ones of said plurality of spectral vectors. 3. The method of claim 1 wherein said computing spectral similarity values includes computing a magnitude difference between a first and a second of said mean spectral vectors. 4. The method of claim 3 wherein said computing spectral similarity values includes computing a shape difference between said first and said second mean spectral vectors. 5. The method of claim 3 wherein said property corresponds to reflectance, and wherein said determining a magnitude difference includes:computing a squared differential reflectance magnitude between said first mean spectral vector and said second mean spectral vector with respect to a number (N) of said spectral bands;summing said squared differential reflectance magnitudes; anddividing the sum of said squared differential reflectance magnitudes by N. 6. The method of claim 5 wherein said determining a magnitude difference includes evaluating the following expression over a number (Nb) of said spectral bands:wherein d e represents said magnitude difference, x i represents the value of the first mean spectral vector in the i th of said spectral bands, and wherein y i represents the value of the second mean spectral vector in the i th of said spectral bands. 7. The method of claim 4 wherein said determining a shape difference includes evaluating the following expression over a number (Nb) of said spectral bands:wherein r 2 is representative of said shape difference, x i represents the value of the first mean spectral vector in the i th of said spectral bands, y i represents the value of the second mean spectral vector in the i th of said spectral bands, μ x represents the mean value of the first spectral vector vector, and μ y represents the means value of the second mean spectral vector, and wherein σ x represents the standard deviation of first mean spectral vector and wherein σ y represents the standard deviation of the second mean spectral vector. 8. A method for evaluating quality of an image, comprising:receiving a spectral image;organizing pixels from the spectral image into a plurality of classes by categorizing spectral vectors characterizing said pixels;determining a mean spectral vector for each of said plurality of classes, each said mean spectral vector being determined using ones of said spectral vectors associated with one of said plurality of classes;computing spectral similarities between pairs of said mean spectral vectors; andanalyzing said spectral similarities in order to obtain information relevant to said quality of said image. 9. The method of claim 8 wherein said organizing includes computing spectral similarity values between a seed one of said pixels and each of a set of unprocessed ones of said pixels. 10. The method of claim 8 wherein said computing includes determining magnitude differences and shape differences between said pairs of said mean spectral vectors. 11. An image processing system comprising:an input interface through which is received a spectral image;a computer readable medium having stored therein an image quality assessment stored program; anda processor operative to execute said image quality assessment stored program and thereby:(i) organize pixels from the spectral image into a plurality of classes by categorizing spectral vectors characterizing said pixels,(ii) determine a mean spectral vector for each of said plurality of classes, each said mean spectral vector being determined using ones of said spectral vectors associated with one of said plurality of classes,(iii) compute spectral similarities between pairs of said mean spectral vectors, and(iv) analyze said spectral similarities in order to obtain information relevant to quality of said image. 12. The system of claim 11 wherein said processor is further operative to compute spectral similarity values between a seed one of said pixels and each of a set of unprocessed ones of said pixels in order to facilitate organization of said pixels into said plurality of classes. 13. The system of claim 11 wherein said processor is further operative to compute said spectral similarities based upon magnitude differences and shape differences between said pairs of said mean spectral vectors. 14. An article of manufacture for use with a data processing system, comprising a computer readable medium having stored therein an image quality assessment stored program, said data processing system being configured by said image quality assessment stored program when executed by said data processing system to:organize pixels from a spectral image into a plurality of classes by categorizing spectral vectors characterizing said pixels;determine a mean spectral vector for each of said plurality of classes, each said mean spectral vector being determined using ones of said spectral vectors associated with one of said plurality of classes;compute spectral similarities between pairs of said mean spectral vectors; andanalyze said spectral similarities in order to obtain information relevant to quality of said spectral image. 15. The article of manufacture of claim 14 wherein said image quality assessment stored program, when executed by said data processing system, is further operative to compute said spectral similarities based upon magnitude differences and shape differences between said pairs of said mean spectral vectors. 16. An image processing system comprising:an input interface through which is received a spectral image;a computer readable medium having stored therein an image quality assessment stored program; anda processor operative to execute said image quality assessment stored program and thereby:(i) extract a plurality of input pixels from the spectral image,(ii) converting the plurality of input pixels into a plurality of spectral vectors, each element in each of said spectral vectors representing a reflectance of a respective one of a plurality of spectral bands,(iii) organize said plurality of spectral vectors into a set of M classes;(iv) determine a mean reflectance vector associated with each of said M classes,(v) compute spectral similarities between pairs of said mean reflectance vectors, and(vi) analyze said spectral similarities in order to obtain information relevant to quality of said spectral image. 17. The image processing system of claim 16 wherein said processor is further operative to compute spectral similarity between a first of said mean spectral vectors and a second of said mean spectral vectors based on a magnitude difference and a shape difference therebetween. 18. The image processing system of claim 17 wherein said processor is further operative to determining said magnitude difference by:computing a squared differential reflectance magnitude between said first mean spectral vector and said second mean spectral vector with respect to a number (N) of said spectral bands;summing said squared differential reflectance magnitudes; anddividing the sum of said squared differential reflectance magnitudes by N. 19. The image processing system of claim 16 wherei n said processor is further operative to compute spectral similarity values between a seed one of said spectral vectors and each of a set of unprocessed ones of said spectral vectors in order to facilitate organization of said spectral vectors into said M classes.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (8)
Nella John ; Austin Joseph W. ; Livingston Peter M., Hyperspectral air-to-air missile seeker.
Hollinger, Allan B.; Qian, Shen-en; Williams, Daniel J., System for interactive visualization and analysis of imaging spectrometry datasets over a wide-area network.
Simmons, Rulon Eugene; Elder, Timothy DeWald; Stewart, David Jonathan; Cincotta, Eric John; Kenndey, Carolyn A.; Van Nostrand, Richard Craig, Imaging system analysis methods and apparatus.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.